一种新的多智能体强化学习算法及其在多机器人协作任务中的应用

在多机器人系统中 ,评价一个机器人行为的好坏常常依赖于其它机器人的行为 ,此时必须采用组合动作以实现多机器人的协作 ,但采用组合动作的强化学习算法由于学习空间异常庞大而收敛得极慢 .本文提出的新方法通过预测各机器人执行动作的概率来降低学习空间的维数 ,并应用于多机器人协作任务之中 .实验结果表明 ,基于预测的加速强化学习算法可以比原始算法更快地获得多机器人的协作策略 .

基于立体视觉的室外非结构化环境障碍物检测方法研究

随着智能机器人系统的发展，机器人的在线感知能力日益受到重视。障碍物检测能力是机器人在线感知能力的一个重要组成部分。因视觉传感器具有独特优势，基于视觉的障碍物检测方法成为目前关注的重点。 室外非结构化环境因结构复杂，机器人缺乏可有效利用的先验知识描述，导致众多障碍物检测系统在该环境中不能有效工作。本文采用全局-局部策略对场景进行由粗到精的分析，弥补室外非结构化环境先验知识不足的难题，提高机器人的在线感知能力。根据该策略，本文在基于视差图的障碍物检测系统框架中，引入视差投影图模块，提出了基于视差投影图的障碍物检测系统框架。该框架在视差投影图模块中全局分析场景视差分布水平，在立体匹配模块中局部分析场景前景目标的几何轮廓信息。依据该框架，针对实际应用中遇到的各种问题，提出了工作于室外非结构化环境的障碍物检测算法。该算法具有如下特点： 1、通过分析视差投影图的地面关联线信息，获得场景的视差分布水平。该信息一方面用来动态更改匹配算法的视差搜索范围，增强算法的实时性和鲁棒性；另一方面用来移除背景地表，简化障碍物分割过程； 2、采用双域滤波抑制噪声，获得清晰的边缘特征，降低立体匹配算法在深度不连续性区域的匹配难度； 3、借助逆向重投影的思想重采样扫描图像，在立体匹配前等效地实现了立体匹配过程中动态变更视差搜索范围的操作； 4、采用基于连通成分的扩散方法替代传统的SAD局部匹配算法，获得高质量的视差图，最终改善障碍物检测的精确性。 在室外非结构化环境中，本文对该算法进行了实验验证。通过设置不同的基线长度，验证了算法在不同的感知距离内的有效性。经实验证明，本算法在一定距离范围内能够有效的检测出障碍物。

混沌粒子群混合算法对微型永磁电机的优化设计

在电机的设计中,常常需要通过优化设计得到合理的电机结构尺寸和参数.电机的设计问题实质上是一种带约束的复杂的非线性连续函数优化问题.要得到一个满意的优化结果不仅要求算法具有较高的精度,而且要有快的收敛速度.提出一种新的混合算法对永磁电机的尺寸和整体结构进行优化设计.将混沌算法和粒子群算法相结合,以微型永磁电机为例,对槽形等多个变量进行优化,结果证明了算法的有效性和快速性,适合于同类问题求解.

基于信道估计的RS纠错编码改进算法的设计与实现

提出了一种基于信道估计的RS纠错编码改进算法,该算法可以自适应地根据外界条件和环境对传输信道的干扰变化实时地调节编码系统的数据冗余量。仿真与完整的分析结果证实了该改进算法有效地改善了RS编码算法的传输效率;并且通过实际应用表明:良好的性能,高容错性适应于该通信系统的多种传输信道,具有很强的实用性。

一种基于优化算法的机械手运动学逆解

本文给出一种基于优化算法的机械手运动学逆解的方法 ,这种优化方法基于信赖域方法 ,,具有超线性的收敛速率 .这种方法不仅具有牛顿方法的快速收敛性 ,又具有理想的总体收敛性 .这种方法较 CCD& BFS有明显的优点 ,可以在一般的 PC机上实现实时求解 .在 P II40 0上仅需不到 10 ms就可以求得最优解 .

基于视觉的水下机器人悬停定位方法与应用研究

为解决受外力扰动而影响定点作业的问题，水下机器人需要具有悬停定位(Station keeping)的功能。悬停定位是指在存在外界扰动的情况下，机器人相对于作业目标仍然能够保持期望的位置和姿态。它具有两个特点：首先，是一种动力定位，即在状态感知系统的引导下，依靠自身动力抵抗外界扰动而保持期望位姿；其二，是一种局部定位，机器人相对作业目标的空间运动范围一般不会很大。因此，对悬停定位的研究需要重点解决两个问题，一个是状态感知问题；另一个是控制问题。 传统的声学定位难以满足精度和实时性的要求，视觉是一种实现水下机器人悬停定位的重要方法。本文以国家“十五”863计划重大专项“7000米载人潜水器”的悬停定位为应用背景，从摄像机标定、特定水下目标识别、视觉伺服和演示实验四个方面进行基于视觉的水下机器人悬停定位方法与应用研究。所开发的系统采用特定观察目标作为合作目标，应用基于模型的单目位姿估计方法获取摄像机与观察目标之间的相对位姿信息，实现了悬停定位中的状态感知；然后以视觉系统提供的位姿信息为反馈,构建视觉伺服控制器，实现了水下机器人的闭环控制。 摄像机标定与水下特定目标识别是实现单目位姿估计的前提，其精度直接影响位姿估计的精度，进而影响整个悬停定位系统的性能。通常摄像机标定是一个先建立成像模型，然后求解模型参数的过程。常用的简化成像模型无法精确描述3维成像空间与2维图像平面之间的关系，并且常规标定方法存在非线性方程优化求解困难的问题。因此，本文提出一种基于虚拟摄像机的无模型标定方法。该方法把摄像机的成像过程作为一个黑盒，通过光电测量方式直接建立3维成像空间与对应2维图像平面之间的映射关系。然后根据该映射关系定义一台虚拟理想摄像机，其透视模型参数可以根据需要任意设定，而不影响最终标定结果。虚拟摄像机的引入使得本方法的应用与常规方法同样方便。实验结果表明，该标定方法可以提高位姿估计精度，特别适用于无法用数学模型精确描述成像过程的系统。 针对系统中应用的特定观察目标，设计了水下图像处理和目标识别算法。对水下图像增强处理以后，应用基于自动阈值和区域生长相结合的方法进行图像分割；然后提取观察目标的图像特征，应用基于模型的目标识别方法实现了水下目标的识别和定位。 以单目位姿估计获得的位姿信息为反馈，构建水下机器人的闭环控制器，以实现悬停定位。这是一个典型的基于位置的视觉伺服问题。针对悬停定位的特点，设计了注视优先原则，在机器人运动过程中合理规划各自由度的控制，应用专家PID控制方法实现了水下机器人视觉伺服。 应用上述研究成果，以水下机器人实验平台为载体，在室内水池搭建了演示实验系统，在国内首次完成了基于视觉的水下机器人悬停定位演示实验。实验结果表明：在悬停定位起始阶段，以观察目标在摄像机视场内为前提，机器人能够在视觉伺服控制下跟踪观察目标，并且使观察目标始终保持在摄像机视场之中；能够运动到指定位置并且保持一个特定姿态，实现了机器人的定位定姿；在受到外界扰动（外力或水流）的情况下，仍然能够恢复到原来的位姿；即使在持续水流的冲击下，也能稳定地保持期望位姿。水下机器人悬停定位演示实验为今后基于视觉的悬停定位技术应用于实际作业打下了良好基础。

成批生产计划调度的集成建模与优化

针对多品种批量生产类型,建立了调度约束的生产计划与调度集成优化模型。模型的目标函数是使总调整费用、库存费用及生产费用之和最小,约束函数包括库存平衡约束和生产能力约束,同时考虑了调度约束,即工序顺序约束和工件在单机上的加工能力约束,保证了计划可行性。该模型为两层混合整数规划模型,对其求解综合运用了遗传算法和启发式规则,提出了混合启发式求解算法。最后,针对某机床厂多品种批量生产类型车间进行了实例应用,对车间零件月份作业计划进行分解,得到各工段单元零件周作业计划,确定了零件各周生产批量与投产顺序。

叠前时间偏移：各向异性、多次波衰减及应用

At present, in order to image complex structures more accurately, the seismic migration methods has been developed from isotropic media to the anisotropic media. This dissertation develops a prestack time migration algorithm and application aspects for complex structures systematically. In transversely isotropic media with a vertical symmetry axis (VTI media), the dissertation starts from the theory that the prestack time migration is an approximation of the prestack depth migration, based on the one way wave equation and VTI time migration dispersion relation, by combining the stationary-phase theory gives a wave equation based VTI prestack time migration algorithm. Based on this algorithm, we can analytically obtain the travel time and amplitude expression in VTI media, as while conclude how the anisotropic parameter influence the time migration, and by analyzing the normal moveout of the far offset seismic data and lateral inhomogeneity of velocity, we can update the velocity model and estimate the anisotropic parameter model through the time migration. When anisotropic parameter is zero, this algorithm degenerates to the isotropic time migration algorithm naturally, so we can propose an isotopic processing procedure for imaging. This procedure may keep the main character of time migration such as high computational efficiency and velocity estimation through the migration, and, additionally, partially compensate the geometric divergence by adopting the deconvolution imaging condition of wave equation migration. Application of this algorithm to the complicated synthetic dataset and field data demonstrates the effectiveness of the approach. In the dissertation we also present an approach for estimating the velocity model and anisotropic parameter model. After analyzing the velocity and anisotropic parameter impaction on the time migration, and based on the normal moveout of the far offset seismic data and lateral inhomogeneity of velocity, through migration we can update the velocity model and estimate the anisotropic parameter model by combining the advantages of velocity analysis in isotropic media and anisotropic parameter estimation in VTI media. Testing on the synthetic and field data, demonstrates the method is effective and very steady. Massive synthetic dataset、2D sea dataset and 3D field datasets are used for VTI prestack time migration and compared to the stacked section after NMO and prestack isotropic time migration stacked section to demonstrate that VTI prestack time migration method in this paper can obtain better focusing and less positioning errors of complicated dip reflectors. When subsurface is more complex, primaries and multiples could not be separated in the Radon domain because they can no longer be described with simple functions (parabolic). We propose an attenuating multiple method in the image domain to resolve this problem. For a given velocity model，since time migration takes the complex structures wavefield propagation in to account, primaries and multiples have different offset-domain moveout discrepancies, then can be separated using techniques similar to the prior migration with Radon transform. Since every individual offset-domain common-reflection point gather incorporates complex 3D propagation effects, our method has the advantage of working with 3D data and complicated geology. Testing on synthetic and real data, we demonstrate the power of the method in discriminating between primaries and multiples after prestack time migration, and multiples can be attenuated in the image space considerably.

A novel vision chip for high-speed target tracking

This paper presents a novel vision chip for high-speed target tracking. Two concise algorithms for high-speed target tracking are developed. The algorithms include some basic operations that can be used to process the real-time image information during target tracking. The vision chip is implemented that is based on the algorithms and a row-parallel architecture. A prototype chip has 64 x 64 pixels is fabricated by 0.35 pm complementary metal-oxide-semiconductor transistor (CMOS) process with 4.5 x 2.5 mm(2) area. It operates at a rate of 1000 frames per second with 10 MHz chip main clock. The experiment results demonstrate that a high-speed target can be tracked in complex static background and a high-speed target among other high-speed objects can be tracked in clean background.

A 1,000 Frames/s Programmable Vision Chip with Variable Resolution and Row-Pixel-Mixed Parallel Image Processors

A programmable vision chip with variable resolution and row-pixel-mixed parallel image processors is presented. The chip consists of a CMOS sensor array, with row-parallel 6-bit Algorithmic ADCs, row-parallel gray-scale image processors, pixel-parallel SIMD Processing Element (PE) array, and instruction controller. The resolution of the image in the chip is variable: high resolution for a focused area and low resolution for general view. It implements gray-scale and binary mathematical morphology algorithms in series to carry out low-level and mid-level image processing and sends out features of the image for various applications. It can perform image processing at over 1,000 frames/s (fps). A prototype chip with 64 x 64 pixels resolution and 6-bit gray-scale image is fabricated in 0.18 mu m Standard CMOS process. The area size of chip is 1.5 mm x 3.5 mm. Each pixel size is 9.5 mu m x 9.5 mu m and each processing element size is 23 mu m x 29 mu m. The experiment results demonstrate that the chip can perform low-level and mid-level image processing and it can be applied in the real-time vision applications, such as high speed target tracking.

Efficient two-step digit-set-restricted modified signed-digit algorithm based on optoelectronic shared content-addressable memory

A two-step digit-set-restricted modified signed-digit (MSD) adder based on symbolic substitution is presented. In the proposed addition algorithm, carry propagation is avoided by using reference digits to restrict the intermediate MSD carry and sum digits into {(1) over bar ,0} and {0, 1}, respectively. The algorithm requires only 12 minterms to generate the final results, and no complementarity operations for nonzero outputs are involved, which simplifies the system complexity significantly. An optoelectronic shared content-addressable memory based on an incoherent correlator is used for experimental demonstration. (c) 2005 Society of Photo-Optical Instrumentation Engineers.

Efficient two-step digit-set-restricted modified signed-digit algorithm based on optoelectronic shared content-addressable memory

A two-step digit-set-restricted modified signed-digit (MSD) adder based on symbolic substitution is presented. In the proposed addition algorithm, carry propagation is avoided by using reference digits to restrict the intermediate MSD carry and sum digits into {(1) over bar ,0} and {0, 1}, respectively. The algorithm requires only 12 minterms to generate the final results, and no complementarity operations for nonzero outputs are involved, which simplifies the system complexity significantly. An optoelectronic shared content-addressable memory based on an incoherent correlator is used for experimental demonstration. (c) 2005 Society of Photo-Optical Instrumentation Engineers.

A reliable phase unwrapping algorithm based on the local fitting plane and quality map

A fast and reliable phase unwrapping (PhU) algorithm, based on the local quality-guided fitting plane, is presented. Its framework depends on the basic plane-approximated assumption for phase values of local pixels and on the phase derivative variance (PDV) quality map. Compared with other existing popular unwrapping algorithms, the proposed algorithm demonstrated improved robustness and immunity to strong noise and high phase variations, given that the plane assumption for local phase is reasonably satisfied. Its effectiveness is demonstrated by computer-simulated and experimental results.

Weighted least-squares phase unwrapping algorithm based on derivative variance correlation map

Among different phase unwrapping approaches, the weighted least-squares minimization methods are gaining attention. In these algorithms, weighting coefficient is generated from a quality map. The intrinsic drawbacks of existing quality maps constrain the application of these algorithms. They often fail to handle wrapped phase data contains error sources, such as phase discontinuities, noise and undersampling. In order to deal with those intractable wrapped phase data, a new weighted least-squares phase unwrapping algorithm based on derivative variance correlation map is proposed. In the algorithm, derivative variance correlation map, a novel quality map, can truly reflect wrapped phase quality, ensuring a more reliable unwrapped result. The definition of the derivative variance correlation map and the principle of the proposed algorithm are present in detail. The performance of the new algorithm has been tested by use of a simulated spherical surface wrapped data and an experimental interferometric synthetic aperture radar (IFSAR) wrapped data. Computer simulation and experimental results have verified that the proposed algorithm can work effectively even when a wrapped phase map contains intractable error sources. (c) 2006 Elsevier GmbH. All rights reserved.

Discussions about FFT-based two-step phase-shifting algorithm

An FFT-based two-step phase-shifting (TPS) algorithm is described in detail and implemented by use of experimental interferograms. This algorithm has been proposed to solve the TPS problem with random phase shift except pi. By comparison with the visibility-function-based TPS algorithm, it proves that the FFT-based algorithm has obvious advantages in phase extracting. Meanwhile, we present a pi-phase-shift supplement to the TPS algorithm, which combines the two interferograms and demodulates the phase map by locating the extrema of the combined fringes after removing the respective backgrounds. So combining this method and FFT-based one, one could really implement the TPS with random phase shift. Whereafter, we systematically compare the TPS with single-interferogram analysis algorithm and conventional three-step phase-shifting one. The results demonstrate that the FFT-based TPS algorithm has a satisfactory accuracy. At last, based on the polarizing interferometry, a schematic setup of two-channel TPS interferometer with random phase shift is suggested to implement the simultaneous collection of interferograms. (c) 2007 Elsevier GrnbH. All rights reserved.